Back

Diagnostic accuracy of MRI in diagnosing Cardiac Sarcoidosis - A Meta-analysis

Sabuwala, N. M.; Raval, D.; Parikh, H.; Saiyed, F.; Desai, D.

2023-05-04 rheumatology
10.1101/2023.04.29.23289313 medRxiv
Show abstract

IntroductionA multisystem inflammatory illness-sarcoidosis is characterized by organ failure, noncaseating granuloma development, and inflammation. The most frequently affected tissues are the lungs and pulmonary lymph nodes, although other tissues can also be affected including the heart. Cardiac Sarcoidosis can be lethal and result in conduction abnormalities, arrhythmias, and sudden cardiac death. As endocardial biopsy is invasive, Advanced imaging techniques, such as cardiac magnetic resonance (CMR) could be used to diagnose cardiac sarcoidosis and evaluate prognosis. Small and focal cardiac abnormalities have been successfully identified by (DE)-MRI. This comprehensive study and meta-analysis was carried out to aid in the CS diagnosis. MethodologyUsing a strategy based on the search terms-sarcoidosis and CMR/MRI separately, we searched-Pubmed, Google Scholar, Embose, Cochrane Library. Studies were disqualified if they lacked enough data to fill out a 2*2 contingency table. We strictly adhered to the Japanese criteria for diagnosing cases. Data from 12 studies with a total of 785 cases with TP of 212 was included. ResultsWe found the cardiac MRI Sensitivity= 0.934 (95% CI = 0.904 to 0.964), Specificity = 0.875 (95% CI = 0.826 to 0.923), PPV= 0.752 (95% CI = 0.682 to 0.822), Younden index= 0.808 and AUC (Area under curve) for ROC plot= 0.904 with a Diagnostic accuracy of 0.892 for detecting sarcoidosis. ConclusionCardiac MRI is a good and reliable screening and diagnostic tool that can be employed as a non-interventional method for diagnosis of Cardiac Sarcoidosis and future prognosis. Diagnostic accuracy of cardiac sarcoidosis using MRI, a meta-analysis.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.